As a minimum requirement, the total number of photoelectrons measured by the CHEOPS detector in the visible range (from 400 to 1100 nm) should correspond to source photon noise of 150 ppm/min for a V ~ 9 magnitude star. All other sources of noise should account for less than 5% of this error budget. This corresponds to an Earth-size planet of 60 day orbital period transiting a star of 0.9 Rsun detected with a S/Ntransit>10 (100 ppm transit depth). For V ~ 12 magnitude stars (V ~ 13 goal) the source photon noise combined with all other additional sources of noises should be less than 1100 ppm/min. This corresponds to a Neptune-sized planet in an 8 day orbital period transiting a star of 0.7 Rsun detected with a S/Ntransit > 30 (about 2000 ppm transit depth).
The precision needed to reach a S/Ntransit>5 on a 100 ppm transit depth from an Earth-sized planet orbiting a G5 star requires the residual of the average systematic noise on 6h duration to be less than 20 ppm. Similarly, to constrain the detailed shape of the in/egress of transiting Neptunes on K dwarfs, the residual noise should be less than 50 ppm in a 20 minute observation sequence.
CHEOPS should be able to cover at least 50% of the whole sky for a minimum total duration of 50 days of observation per year and per target. The observation may be interrupted up to 50% (goal would be 20%) of the satellite orbital duration (Earth eclipse, Sun, etc.). This requirement is related to the first science goal (shallow transit detection). Any interruption during transit decreases accordingly the signal of the transit detection. Additional observations of the event may be required to reach the signal-to-noise ratio goal. Note that for small size transiting objects, the time of the in/egress is too short to be of any use to constrain the planet size through comparison with the precise measurement of the contrast. Missing in/egress (occurring during interruptions) would not affect the performance of the mission to detect the transit and measure the size of small planets.
For observations of transiting hot Neptune planets, the sky coverage does not need to be as complete as for the Doppler targets, since for example NGTS will target only a fraction of the southern sky (since members of the CHEOPS team are also members of the NGTS team, there are possibilities to optimize this fraction within the satellite visibility window). We consider as a minimum requirement that 25% of the sky (2/3 located in the southern hemisphere) should be visible for a cumulative duration of 13 days with interruptions less then 20 minutes.
Individual exposures should be short enough to avoid saturation on V ~ 6 magnitude stars, but the temporal resolution of the measurement should be 1 minute. Full frame images (addition of shorter images when required) will be recorded (and later downloaded) in 1-minute intervals. The time stamp (UTC) uncertainty on the time of exposure should be smaller than 1s.
Transit detection on bright stars identified by Doppler surveys will need about 2 days of continuous pointing on target to cope with uncertainties on radial velocity ephemerids of the longest period planet (about 3-5% of the orbital period). With a minimum of 150 targets and 50% of orbit interruptions, this corresponds to a minimum total of 600 days of satellite life.
For NGTS targets a shorter on-target time is required (12 hours). If one considers 50 targets with a single transit observation and 50 additional targets where 4 transits will be observed and 5 targets where 10 transits may be required we end up with 150 days. With 20% efficiency correction this leads to 180 days of satellite life.
Observations to detect the planets directly in reflected light will be possible for a handful of hot Jupiters. To obtain a reliable measurement, disentangled from possible stellar photometric variability, observations of 3 full planetary orbits are needed. Assuming a typical 5 days orbital period for hot Jupiter, 15 days of continuous observation are required. Estimating a sample of 5 hot Jupiters for which these observations are required, this corresponds to 75 days.
In total these three programs combined require 635 separate target pointings. Assuming 0.3 hour per pointing, 80% efficiency in the scheduling and instrument monitoring, the mission duration is estimated at ~1100 days or 3 years.
Adding to this duration the open time allocation for carrying out ancillary science (up to 20%), the total duration of the CHEOPS mission is estimated to be 3.5 years.